[-] babelspace@kbin.social 2 points 1 year ago

Since any reductions to copyright, if they occur at all, will take a while to happen, I hope someone comes up with an opt-in limited term copyright. At max, I'd be satisfied with a 45-50 year limited copyright on everything I make, and could see going shorter under plenty of circumstances.

[-] babelspace@kbin.social 10 points 1 year ago

I wish I could get through to people who fear AI copyright infringement on this point.

[-] babelspace@kbin.social 3 points 1 year ago

Interesting. I've seen plenty of mentions of the potential use of AI in medicinal imaging, but not in a context where time is of the essence. That's a particularly compelling use case.

"Still needs to be greenlit by the FDA" - has anyone read a discussion of how the FDA is handing AI medical applications? Considering how overwrought the approval procedure can be for more mundane medical technologies, I have little hope that the process is going to be efficient and effective.

[-] babelspace@kbin.social 4 points 1 year ago

There are circumstances where the precautionary principle is good to apply. But overuse of it has really bad cumulative consequences.

1

I’ve been following the development of the next Stable Diffusion model, and I’ve seen this approach mentioned.

Seems like this is a way in which AI training is analogous to human learning - we learn quite a lot from fiction, games, simulations and apply this to the real world. I’m sure the same pitfalls apply as well.

[-] babelspace@kbin.social 7 points 1 year ago

Thanks for bringing it to our attention. Discovery is difficult right now, advertising in places that are genuinely relevant is a real service.

[-] babelspace@kbin.social 2 points 1 year ago

That's a pretty good summary.

[-] babelspace@kbin.social 3 points 1 year ago* (last edited 1 year ago)

It's just one method to make tea. Some teas taste quite different from cup to cup (wuyi oolongs, for example), some are more consistent, in my experience. What I like is that it's easy to adjust depending on the outcome - one infusion is too weak or too strong? Brew the next one more or less.

2

This is the potential development in AI I'm most interested in. So naturally, I tested this when I first used ChatGPT. In classic ChatGPT fashion, when asked to make a directed acyclic graph representing cause and effect, it could interpret that well enough to make a simple graph...but got the cause and effect flow for something as simple as lighting a fire. Haven't tried it again with ChatGPT-4 though.

[-] babelspace@kbin.social 3 points 1 year ago

I’ve seen the fedi search link before, it’s a great idea. That said, it’s currently no good for finding things on kbin - for example, try searching AskKbin, which has been quite active - you get almost nothing back. Hopefully it will be better in a few days.

9
submitted 1 year ago by babelspace@kbin.social to c/tech@kbin.social

Like most who’ve joined kbin or lemmy, I’m actively looking for more communities to subscribe to. We don’t have a way to do a centralized search for these across instances yet, so I’m sure there’s some I’ve overlooked. What smaller technology related groups have you discovered that could use more publicity?

So far, I like the spirit of Actually Useful AI - auai@programming.dev: “Not allowed - anything else the mods consider “crypto bro”/“AI bro” success porn sigma grindset blogspam” 😆.
Aside from AI, I’d be interested in finding active communities that focus on web development.

[-] babelspace@kbin.social 3 points 1 year ago

I like the idea of this community, and subscribed, but fyi that link doesn't work for me on kbin. This is the right link on here, I believe: https://kbin.social/m/auai@programming.dev - m rather than c.

[-] babelspace@kbin.social 2 points 1 year ago* (last edited 1 year ago)

That would be a pity -personally I think downvotes, for all they can be abused, are a necessary counterpart to upvotes if you’re going to have a voting system at all. “All attention is good attention” doesn’t work well in real life or online.

Now if a change was made to allow individual magazines to choose upvotes+downvotes, only upvotes or none at all, I think there’s a better case for that - let people try different systems for different circumstances,

[-] babelspace@kbin.social 12 points 1 year ago* (last edited 1 year ago)

Short sighted behavior is hardly limited to capitalist enterprises, though.

[-] babelspace@kbin.social 2 points 1 year ago

How is Stable Diffusion connected to bad/short term orientated social media company policy changes? The ChatGPT connection I understand, even if I don’t think it’s the root cause. (Interest rate increases I think are more fundamental, even if AI is cited by companies as an influence on their choices.)

view more: next ›

babelspace

joined 1 year ago